The Combination Of Evidence And Its Appliance

Content: Within the framework of evidence theory, data fusion consists in obtaining a single belief function by the combination of several belief functions resulting form distinct information sources. The most popular rule of combination, called Dempster's rule of combination, it is a useful method for dealing with uncertainty problems. Because there exist shortcomings in its combination rule, it hasn't been successfully applied. Firstly, this paper presents an improvement of combination rule. Not only may the combination result be acceptable if the evidences involved conflict with another one highly and the evidence's conflicting probability is distributed to every proposition according to its average supported degree, but it is also capable of handing related evidence focusing to the amount of information resident in piece of evidence. Experiments show that the new combination rule improves the reliability and rationality of the combination result. Applying fuzzy theory to the combination of evidence, we put forward two kinds of combination method, and clarify its merit and shortcoming. We put the evidence theory and fuzzy theory together and apply them in student comprehensive competence finally.